4.1 KiB

Raw Blame History

Atomic Habits AI - Habit Tracker & Personal Library

Atomic Habits AI is a cross-platform mobile and web application designed to help users build lasting habits through a science-backed tracking system and a personalized AI-powered library. It combines local-first performance with cloud synchronization and AI insights.

🏗 Architecture

The application follows a Local-First, Cloud-Sync architecture.

Frontend

Framework: Expo (React Native) with Expo Router for file-based routing.
Language: TypeScript.
State Management: Local SQLite database for offline-first capabilities, synced with Supabase.
UI/UX: Custom design system built with Vanilla CSS (via React Native Stylesheets), utilizing Lucide React Native for iconography and Reanimated for smooth transitions.

Backend & Cloud

Platform: Supabase
Authentication: Supabase Auth (Email/Password, JWT).
Database: PostgreSQL (Cloud) + SQLite (Local).
Storage: Supabase Storage for PDF/Epub books.
Edge Functions: Supabase Edge Functions (Deno) for AI processing and book parsing.

AI Integration

Models: Gemini (via Supabase Edge Functions).
Features: Automated book metadata extraction, AI-generated reading insights/synthesis, and interactive habit coaching.

🔐 Security

Security is integrated at every layer of the stack:

Authentication: Secure session management using AsyncStorage and Supabase JWTs.
Row Level Security (RLS): Supabase RLS policies ensure users can only access their own habits, books, and logs.
Secure Storage: sensitive files are accessed via Signed URLs with limited expiration windows (1 hour) rather than public links.
Local Encryption: (Optional/Roadmap) Local SQLite data protection.
Environment Variables: All sensitive keys (Supabase URL, Anon Key) are managed via .env files and EXPO_PUBLIC_ prefixes.

🗄 Database Schema

The system uses a mirrored schema between local SQLite and cloud PostgreSQL.

Core Tables:

books: Stores library items (id, user_id, title, author, file_uri, current_page, status).
habits: Core habit tracking logic (id, user_id, name, frequency, streak, last_completed).
reading_logs: Tracks progress over time (id, book_id, duration_seconds, pages_read).
chat_history: Persists AI interactions for context-aware coaching.

🛠 Running the Project

Prerequisites

Node.js (v18+)
Expo Go app (for physical device testing) or Android Studio / Xcode (for emulators).

Installation

# Clone the repository
git clone <repo-url>
cd atomichabitsai

# Install dependencies
npm install

Development

# Start the Expo development server
npx expo start

Press a for Android emulator.
Press i for iOS simulator.
Press w for web.

🐞 Debugging Details

Logging

The application uses structured logs for critical paths:

[PdfReader]: WebView lifecycle and file writing logs.
[downloadBook]: Sync and storage fetch status.
[resolveFileUri]: Path resolution debugging for iOS container changes.

Common Troubleshooting

Stale iOS Paths: If files don't open on iOS after an update, the resolveFileUri utility automatically corrects absolute paths that contain outdated container UUIDs.
Sync Failures: Check the performMutation calls in sync.ts. Ensure the device has internet access for Supabase connectivity.
PDF Reader Crashes: Ensure the library @bildau/rn-pdf-reader is correctly patched via patch-package to handle the Legacy FileSystem API in Expo 54.

🧑‍💻 Backend Developer Notes

Edge Functions

Located in /supabase/functions/:

process-book-ai: Parses uploaded filenames to suggest titles/authors and generates reading summaries.
chat-ai: The core engine for the AI habit coach.

Storage

Bucket name: books
Access: Private (accessed via createSignedUrl in src/lib/file-utils.ts).

4.1 KiB Raw Blame History